Binary sparse nonnegative matrix factorization


This paper presents a fast part-based subspace selection algorithm, termed the binary sparse nonnegative matrix factorization (B-SNMF). Both the training process and the testing process of B-SNMF are much faster than those of binary principal component analysis (B-PCA). Besides, B-SNMF is more robust to occlusions in images. Experimental results on face images demonstrate the effectiveness and the efficiency of the proposed B-SNMF.

Publication DOI:
Divisions: Engineering & Applied Sciences > Computer Science
Life & Health Sciences > Pharmacy
Uncontrolled Keywords: binary principal component analysis,binary sparse nonnegative matrix factorization,face images,fast part-based subspace selection algorithm,image occlusions,Electrical and Electronic Engineering,Media Technology
Full Text Link: http://www.nlpr ... ers/kz/gk13.pdf
Related URLs: http://www.scop ... tnerID=8YFLogxK (Scopus URL)
http://ieeexplo ... rnumber=4801604 (Publisher URL)
PURE Output Type: Article
Published Date: 2009-05
Authors: Yuan, Yuan
Li, Xuelong
Pang, Yanwei
Lu, Xin
Tao, Dacheng

Export / Share Citation


Additional statistics for this record